Belief Flipping in a Dynamic Model of Statistical Discrimination
The literature on statistical discrimination shows that ex-ante identical groups may be differentially treated in discriminatory equilibria. This paper constructs a dynamic model of statistical discrimination and explores what happens to the individuals who nonetheless overcome the initial discrimination. If an employer discriminates against a group of workers in her initial hiring, she may actually favor the successful members of that group when she promotes from within the firm. The worker's welfare implications (i.e. who benefits from an employer's discriminatory hiring practices) are unclear. Even though agents face discrimination initially, some may be better off because of it.
I would like to give special thanks to Tomas J. Sjööströöm and Glenn C. Loury for continued support and advice on this project. I would also like to thank Gary Becker, Kalyan Chatterjee, Kim-Sau Chung, Hanming Fang, Vijay Krishna, Jim Jordan, Xue Jun, Steven Levitt, Brian McCannon, Andrea Moro, Roger Myerson, Peter Norman, Tor Winston, and seminar participants at the Universities of Chicago, Virginia, and Pennsylvania State. Please direct all correspondence to the author at: Harvard Society of Fellows and NBER; Littauer Center; 1875 Cambridge Street; Cambridge, MA 02138. E-mail: email@example.com. The usual caveat applies.